Interesting point in a longer article about the state of science, which probably deserves its own feature (or book). Metrics about declining productivity come from science of science studies that look for novelty in citation trees which, well... 🤷♂️
www.scientificamerican.com/article/amer...
